STOID Mold_Tag(REBVAL *value, REB_MOLD *mold)
{
Append_Byte(mold->series, '<');
Insert_String(mold->series, AT_TAIL, VAL_SERIES(value), VAL_INDEX(value), VAL_LEN(value), 0);
Append_Byte(mold->series, '>');
}
*/ REBSER *Copy_String(REBSER *src, REBCNT index, REBINT length)
/*
** Copies a portion of any string (byte or unicode).
** Will slim the string, if needed.
**
** The index + length must be in range unsigned int 32.
**
***********************************************************************/
{
REBUNI *up;
REBYTE wide = 1;
REBSER *dst;
REBINT n;
if (length < 0) length = src->tail;
// Can it be slimmed down?
if (!BYTE_SIZE(src)) {
up = UNI_SKIP(src, index);
for (n = 0; n < length; n++)
if (up[n] > 0xff) break;
if (n < length) wide = sizeof(REBUNI);
}
dst = Make_Series(length + 1, wide, MKS_NONE);
Insert_String(dst, 0, src, index, length, TRUE);
SERIES_TAIL(dst) = length;
TERM_SEQUENCE(dst);
return dst;
}
*/ REBSER *Emit(REB_MOLD *mold, REBYTE *fmt, ...)
/*
***********************************************************************/
{
va_list args;
REBYTE ender = 0;
REBSER *series = mold->series;
ASSERT2(SERIES_WIDE(series) ==2, 9997);
va_start(args, fmt);
for (; *fmt; fmt++) {
switch (*fmt) {
case 'W': // Word symbol
Append_UTF8(series, Get_Word_Name(va_arg(args, REBVAL*)), -1);
break;
case 'V': // Value
Mold_Value(mold, va_arg(args, REBVAL*), TRUE);
break;
case 'S': // String of bytes
Append_Bytes(series, va_arg(args, REBYTE*));
break;
case 'C': // Char
Append_Byte(series, va_arg(args, REBCNT));
break;
case 'E': // Series (byte or uni)
{
REBSER *src = va_arg(args, REBSER*);
Insert_String(series, SERIES_TAIL(series), src, 0, SERIES_TAIL(src), 0);
}
break;
case 'I': // Integer
Append_Int(series, va_arg(args, REBINT));
break;
case 'i':
Append_Int_Pad(series, va_arg(args, REBINT), -9);
Trim_Tail(mold->series, '0');
break;
case '2': // 2 digit int (for time)
Append_Int_Pad(series, va_arg(args, REBINT), 2);
break;
case 'T': // Type name
Append_UTF8(series, Get_Type_Name(va_arg(args, REBVAL*)), -1);
break;
case 'N': // Symbol name
Append_UTF8(series, Get_Sym_Name(va_arg(args, REBCNT)), -1);
break;
case '+': // Add #[ if mold/all
if (GET_MOPT(mold, MOPT_MOLD_ALL)) {
Append_Bytes(series, "#[");
ender = ']';
}
break;
case 'D': // Datatype symbol: #[type
if (ender) {
Append_UTF8(series, Get_Sym_Name(va_arg(args, REBCNT)), -1);
Append_Byte(series, ' ');
} else va_arg(args, REBCNT); // ignore it
break;
case 'B': // Boot string
Append_Boot_Str(series, va_arg(args, REBINT));
break;
default:
Append_Byte(series, *fmt);
}
}
va_end(args);
if (ender) Append_Byte(series, ender);
return series;
}
*/ void Mold_Value(REB_MOLD *mold, REBVAL *value, REBFLG molded)
/*
** Mold or form any value to string series tail.
**
***********************************************************************/
{
REBYTE buf[60];
REBINT len;
REBSER *ser = mold->series;
CHECK_STACK(&len);
ASSERT2(SERIES_WIDE(mold->series) == sizeof(REBUNI), RP_BAD_SIZE);
ASSERT2(ser, RP_NO_BUFFER);
// Special handling of string series: {
if (ANY_STR(value) && !IS_TAG(value)) {
// Forming a string:
if (!molded) {
Insert_String(ser, -1, VAL_SERIES(value), VAL_INDEX(value), VAL_LEN(value), 0);
return;
}
// Special format for ALL string series when not at head:
if (GET_MOPT(mold, MOPT_MOLD_ALL) && VAL_INDEX(value) != 0) {
Mold_All_String(value, mold);
return;
}
}
switch (VAL_TYPE(value)) {
case REB_NONE:
Emit(mold, "+N", SYM_NONE);
break;
case REB_LOGIC:
// if (!molded || !VAL_LOGIC_WORDS(value) || !GET_MOPT(mold, MOPT_MOLD_ALL))
Emit(mold, "+N", VAL_LOGIC(value) ? SYM_TRUE : SYM_FALSE);
// else
// Mold_Logic(mold, value);
break;
case REB_INTEGER:
len = Emit_Integer(buf, VAL_INT64(value));
goto append;
case REB_DECIMAL:
case REB_PERCENT:
len = Emit_Decimal(buf, VAL_DECIMAL(value), IS_PERCENT(value)?DEC_MOLD_PERCENT:0,
Punctuation[GET_MOPT(mold, MOPT_COMMA_PT) ? PUNCT_COMMA : PUNCT_DOT], mold->digits);
goto append;
case REB_MONEY:
len = Emit_Money(value, buf, mold->opts);
goto append;
case REB_CHAR:
Mold_Uni_Char(ser, VAL_CHAR(value), (REBOOL)molded, (REBOOL)GET_MOPT(mold, MOPT_MOLD_ALL));
break;
case REB_PAIR:
len = Emit_Decimal(buf, VAL_PAIR_X(value), DEC_MOLD_MINIMAL, Punctuation[PUNCT_DOT], mold->digits/2);
Append_Bytes_Len(ser, buf, len);
Append_Byte(ser, 'x');
len = Emit_Decimal(buf, VAL_PAIR_Y(value), DEC_MOLD_MINIMAL, Punctuation[PUNCT_DOT], mold->digits/2);
Append_Bytes_Len(ser, buf, len);
//Emit(mold, "IxI", VAL_PAIR_X(value), VAL_PAIR_Y(value));
break;
case REB_TUPLE:
len = Emit_Tuple(value, buf);
goto append;
case REB_TIME:
//len = Emit_Time(value, buf, Punctuation[GET_MOPT(mold, MOPT_COMMA_PT) ? PUNCT_COMMA : PUNCT_DOT]);
Emit_Time(mold, value);
break;
case REB_DATE:
Emit_Date(mold, value);
break;
case REB_STRING:
// FORM happens in top section.
Mold_String_Series(value, mold);
break;
case REB_BINARY:
if (GET_MOPT(mold, MOPT_MOLD_ALL) && VAL_INDEX(value) != 0) {
Mold_All_String(value, mold);
return;
}
Mold_Binary(value, mold);
break;
case REB_FILE:
if (VAL_LEN(value) == 0) {
Append_Bytes(ser, "%\"\"");
break;
}
Mold_File(value, mold);
break;
case REB_EMAIL:
case REB_URL:
Mold_Url(value, mold);
break;
case REB_TAG:
if (GET_MOPT(mold, MOPT_MOLD_ALL) && VAL_INDEX(value) != 0) {
Mold_All_String(value, mold);
return;
}
Mold_Tag(value, mold);
break;
// Mold_Issue(value, mold);
// break;
case REB_BITSET:
Pre_Mold(value, mold); // #[bitset! or make bitset!
Mold_Bitset(value, mold);
End_Mold(mold);
break;
case REB_IMAGE:
Pre_Mold(value, mold);
if (!GET_MOPT(mold, MOPT_MOLD_ALL)) {
Append_Byte(ser, '[');
Mold_Image_Data(value, mold);
Append_Byte(ser, ']');
End_Mold(mold);
}
else {
REBVAL val = *value;
VAL_INDEX(&val) = 0; // mold all of it
Mold_Image_Data(&val, mold);
Post_Mold(value, mold);
}
break;
case REB_BLOCK:
case REB_PAREN:
if (!molded)
Form_Block_Series(VAL_SERIES(value), VAL_INDEX(value), mold, 0);
else
Mold_Block(value, mold);
break;
case REB_PATH:
case REB_SET_PATH:
case REB_GET_PATH:
case REB_LIT_PATH:
Mold_Block(value, mold);
break;
case REB_VECTOR:
Mold_Vector(value, mold, molded);
break;
case REB_DATATYPE:
if (!molded)
Emit(mold, "N", VAL_DATATYPE(value) + 1);
else
Emit(mold, "+DN", SYM_DATATYPE_TYPE, VAL_DATATYPE(value) + 1);
break;
case REB_TYPESET:
Mold_Typeset(value, mold, molded);
break;
case REB_WORD:
// This is a high frequency function, so it is optimized.
Append_UTF8(ser, Get_Sym_Name(VAL_WORD_SYM(value)), -1);
break;
case REB_SET_WORD:
Emit(mold, "W:", value);
break;
case REB_GET_WORD:
Emit(mold, ":W", value);
break;
case REB_LIT_WORD:
Emit(mold, "\'W", value);
break;
case REB_REFINEMENT:
Emit(mold, "/W", value);
break;
case REB_ISSUE:
Emit(mold, "#W", value);
break;
case REB_CLOSURE:
case REB_FUNCTION:
case REB_NATIVE:
case REB_ACTION:
case REB_COMMAND:
Mold_Function(value, mold);
break;
case REB_OBJECT:
case REB_MODULE:
case REB_PORT:
if (!molded) Form_Object(value, mold);
else Mold_Object(value, mold);
break;
case REB_TASK:
Mold_Object(value, mold); //// | (1<<MOPT_NO_NONE));
break;
case REB_ERROR:
Mold_Error(value, mold, molded);
break;
case REB_MAP:
Mold_Map(value, mold, molded);
break;
case REB_GOB:
{
REBSER *blk;
Pre_Mold(value, mold);
blk = Gob_To_Block(VAL_GOB(value));
Mold_Block_Series(mold, blk, 0, 0);
End_Mold(mold);
}
break;
case REB_EVENT:
Mold_Event(value, mold);
break;
case REB_REBCODE:
case REB_OP:
case REB_FRAME:
case REB_HANDLE:
case REB_STRUCT:
case REB_LIBRARY:
case REB_UTYPE:
// Value has no printable form, so just print its name.
if (!molded) Emit(mold, "?T?", value);
else Emit(mold, "+T", value);
break;
case REB_END:
case REB_UNSET:
if (molded) Emit(mold, "+T", value);
break;
default:
Crash(RP_DATATYPE+5, VAL_TYPE(value));
}
return;
append:
Append_Bytes_Len(ser, buf, len);
}
x*/ void Modify_StringX(REBCNT action, REBVAL *string, REBVAL *arg)
/*
** Actions: INSERT, APPEND, CHANGE
**
** string [string!] {Series at point to insert}
** value [any-type!] {The value to insert}
** /part {Limits to a given length or position.}
** length [number! series! pair!]
** /only {Inserts a series as a series.}
** /dup {Duplicates the insert a specified number of times.}
** count [number! pair!]
**
***********************************************************************/
{
REBSER *series = VAL_SERIES(string);
REBCNT index = VAL_INDEX(string);
REBCNT tail = VAL_TAIL(string);
REBINT rlen; // length to be removed
REBINT ilen = 1; // length to be inserted
REBINT cnt = 1; // DUP count
REBINT size;
REBVAL *val;
REBSER *arg_ser = 0; // argument series
// Length of target (may modify index): (arg can be anything)
rlen = Partial1((action == A_CHANGE) ? string : arg, DS_ARG(AN_LENGTH));
index = VAL_INDEX(string);
if (action == A_APPEND || index > tail) index = tail;
// If the arg is not a string, then we need to create a string:
if (IS_BINARY(string)) {
if (IS_INTEGER(arg)) {
if (VAL_INT64(arg) > 255 || VAL_INT64(arg) < 0)
Trap_Range(arg);
arg_ser = Make_Binary(1);
Append_Byte(arg_ser, VAL_CHAR(arg)); // check for size!!!
}
else if (!ANY_BINSTR(arg)) Trap_Arg(arg);
}
else if (IS_BLOCK(arg)) {
// MOVE!
REB_MOLD mo = {0};
arg_ser = mo.series = Make_Unicode(VAL_BLK_LEN(arg) * 10); // GC!?
for (val = VAL_BLK_DATA(arg); NOT_END(val); val++)
Mold_Value(&mo, val, 0);
}
else if (IS_CHAR(arg)) {
// Optimize this case !!!
arg_ser = Make_Unicode(1);
Append_Byte(arg_ser, VAL_CHAR(arg));
}
else if (!ANY_STR(arg) || IS_TAG(arg)) {
arg_ser = Copy_Form_Value(arg, 0);
}
if (arg_ser) Set_String(arg, arg_ser);
else arg_ser = VAL_SERIES(arg);
// Length of insertion:
ilen = (action != A_CHANGE && DS_REF(AN_PART)) ? rlen : VAL_LEN(arg);
// If Source == Destination we need to prevent possible conflicts.
// Clone the argument just to be safe.
// (Note: It may be possible to optimize special cases like append !!)
if (series == VAL_SERIES(arg)) {
arg_ser = Copy_Series_Part(arg_ser, VAL_INDEX(arg), ilen); // GC!?
}
// Get /DUP count:
if (DS_REF(AN_DUP)) {
cnt = Int32(DS_ARG(AN_COUNT));
if (cnt <= 0) return; // no changes
}
// Total to insert:
size = cnt * ilen;
if (action != A_CHANGE) {
// Always expand series for INSERT and APPEND actions:
Expand_Series(series, index, size);
} else {
if (size > rlen)
Expand_Series(series, index, size-rlen);
else if (size < rlen && DS_REF(AN_PART))
Remove_Series(series, index, rlen-size);
else if (size + index > tail) {
EXPAND_SERIES_TAIL(series, size - (tail - index));
}
}
// For dup count:
for (; cnt > 0; cnt--) {
Insert_String(series, index, arg_ser, VAL_INDEX(arg), ilen, TRUE);
index += ilen;
}
TERM_SERIES(series);
VAL_INDEX(string) = (action == A_APPEND) ? 0 : index;
}
//
// Modify_String: C
//
// Returns new dst_idx.
//
REBCNT Modify_String(
REBCNT action, // INSERT, APPEND, CHANGE
REBSER *dst_ser, // target
REBCNT dst_idx, // position
const REBVAL *src_val, // source
REBFLGS flags, // AN_PART
REBINT dst_len, // length to remove
REBINT dups // dup count
) {
REBSER *src_ser = 0;
REBCNT src_idx = 0;
REBCNT src_len;
REBCNT tail = SER_LEN(dst_ser);
REBINT size; // total to insert
REBOOL needs_free;
REBINT limit;
// For INSERT/PART and APPEND/PART
if (action != SYM_CHANGE && GET_FLAG(flags, AN_PART))
limit = dst_len; // should be non-negative
else
limit = -1;
if (limit == 0 || dups < 0) return (action == SYM_APPEND) ? 0 : dst_idx;
if (action == SYM_APPEND || dst_idx > tail) dst_idx = tail;
// If the src_val is not a string, then we need to create a string:
if (GET_FLAG(flags, AN_SERIES)) { // used to indicate a BINARY series
if (IS_INTEGER(src_val)) {
src_ser = Make_Series_Codepoint(Int8u(src_val));
needs_free = TRUE;
limit = -1;
}
else if (IS_BLOCK(src_val)) {
src_ser = Join_Binary(src_val, limit); // NOTE: it's the shared FORM buffer!
needs_free = FALSE;
limit = -1;
}
else if (IS_CHAR(src_val)) {
//
// "UTF-8 was originally specified to allow codepoints with up to
// 31 bits (or 6 bytes). But with RFC3629, this was reduced to 4
// bytes max. to be more compatible to UTF-16." So depending on
// which RFC you consider "the UTF-8", max size is either 4 or 6.
//
src_ser = Make_Binary(6);
SET_SERIES_LEN(
src_ser,
Encode_UTF8_Char(BIN_HEAD(src_ser), VAL_CHAR(src_val))
);
needs_free = TRUE;
limit = -1;
}
else if (ANY_STRING(src_val)) {
src_len = VAL_LEN_AT(src_val);
if (limit >= 0 && src_len > cast(REBCNT, limit))
src_len = limit;
src_ser = Make_UTF8_From_Any_String(src_val, src_len, 0);
needs_free = TRUE;
limit = -1;
}
else if (!IS_BINARY(src_val))
fail (Error_Invalid_Arg(src_val));
}
else if (IS_CHAR(src_val)) {
src_ser = Make_Series_Codepoint(VAL_CHAR(src_val));
needs_free = TRUE;
}
else if (IS_BLOCK(src_val)) {
src_ser = Form_Tight_Block(src_val);
needs_free = TRUE;
}
else if (!ANY_STRING(src_val) || IS_TAG(src_val)) {
src_ser = Copy_Form_Value(src_val, 0);
needs_free = TRUE;
}
// Use either new src or the one that was passed:
if (src_ser) {
src_len = SER_LEN(src_ser);
}
else {
src_ser = VAL_SERIES(src_val);
src_idx = VAL_INDEX(src_val);
src_len = VAL_LEN_AT(src_val);
needs_free = FALSE;
}
if (limit >= 0) src_len = limit;
// If Source == Destination we need to prevent possible conflicts.
// Clone the argument just to be safe.
// (Note: It may be possible to optimize special cases like append !!)
if (dst_ser == src_ser) {
assert(!needs_free);
src_ser = Copy_Sequence_At_Len(src_ser, src_idx, src_len);
needs_free = TRUE;
src_idx = 0;
}
// Total to insert:
size = dups * src_len;
if (action != SYM_CHANGE) {
// Always expand dst_ser for INSERT and APPEND actions:
Expand_Series(dst_ser, dst_idx, size);
} else {
if (size > dst_len)
Expand_Series(dst_ser, dst_idx, size - dst_len);
else if (size < dst_len && GET_FLAG(flags, AN_PART))
Remove_Series(dst_ser, dst_idx, dst_len - size);
else if (size + dst_idx > tail) {
EXPAND_SERIES_TAIL(dst_ser, size - (tail - dst_idx));
}
}
// For dup count:
for (; dups > 0; dups--) {
Insert_String(dst_ser, dst_idx, src_ser, src_idx, src_len, TRUE);
dst_idx += src_len;
}
TERM_SEQUENCE(dst_ser);
if (needs_free) {
// If we did not use the series that was passed in, but rather
// created an internal temporary one, we need to free it.
Free_Series(src_ser);
}
return (action == SYM_APPEND) ? 0 : dst_idx;
}
*/ REBCNT Modify_String(REBCNT action, REBSER *dst_ser, REBCNT dst_idx, const REBVAL *src_val, REBCNT flags, REBINT dst_len, REBINT dups)
/*
** action: INSERT, APPEND, CHANGE
**
** dst_ser: target
** dst_idx: position
** src_val: source
** flags: AN_PART
** dst_len: length to remove
** dups: dup count
**
** return: new dst_idx
**
***********************************************************************/
{
REBSER *src_ser = 0;
REBCNT src_idx = 0;
REBCNT src_len;
REBCNT tail = SERIES_TAIL(dst_ser);
REBINT size; // total to insert
if (dups < 0) return (action == A_APPEND) ? 0 : dst_idx;
if (action == A_APPEND || dst_idx > tail) dst_idx = tail;
// If the src_val is not a string, then we need to create a string:
if (GET_FLAG(flags, AN_SERIES)) { // used to indicate a BINARY series
if (IS_INTEGER(src_val)) {
src_ser = Append_Byte(0, Int8u(src_val)); // creates a binary
}
else if (IS_BLOCK(src_val)) {
src_ser = Join_Binary(src_val); // NOTE: it's the shared FORM buffer!
}
else if (IS_CHAR(src_val)) {
src_ser = Make_Binary(6); // (I hate unicode)
src_ser->tail = Encode_UTF8_Char(BIN_HEAD(src_ser), VAL_CHAR(src_val));
}
else if (!ANY_BINSTR(src_val)) Trap_Arg_DEAD_END(src_val);
}
else if (IS_CHAR(src_val)) {
src_ser = Append_Byte(0, VAL_CHAR(src_val)); // unicode ok too
}
else if (IS_BLOCK(src_val)) {
src_ser = Form_Tight_Block(src_val);
}
else if (!ANY_STR(src_val) || IS_TAG(src_val)) {
src_ser = Copy_Form_Value(src_val, 0);
}
// Use either new src or the one that was passed:
if (src_ser) {
src_len = SERIES_TAIL(src_ser);
}
else {
src_ser = VAL_SERIES(src_val);
src_idx = VAL_INDEX(src_val);
src_len = VAL_LEN(src_val);
}
// For INSERT or APPEND with /PART use the dst_len not src_len:
if (action != A_CHANGE && GET_FLAG(flags, AN_PART)) src_len = dst_len;
// If Source == Destination we need to prevent possible conflicts.
// Clone the argument just to be safe.
// (Note: It may be possible to optimize special cases like append !!)
if (dst_ser == src_ser) {
src_ser = Copy_Series_Part(src_ser, src_idx, src_len);
src_idx = 0;
}
// Total to insert:
size = dups * src_len;
if (action != A_CHANGE) {
// Always expand dst_ser for INSERT and APPEND actions:
Expand_Series(dst_ser, dst_idx, size);
} else {
if (size > dst_len)
Expand_Series(dst_ser, dst_idx, size - dst_len);
else if (size < dst_len && GET_FLAG(flags, AN_PART))
Remove_Series(dst_ser, dst_idx, dst_len - size);
else if (size + dst_idx > tail) {
EXPAND_SERIES_TAIL(dst_ser, size - (tail - dst_idx));
}
}
// For dup count:
for (; dups > 0; dups--) {
Insert_String(dst_ser, dst_idx, src_ser, src_idx, src_len, TRUE);
dst_idx += src_len;
}
TERM_SERIES(dst_ser);
return (action == A_APPEND) ? 0 : dst_idx;
}